CN103532172B - Multistage reserve coordination method based on set dynamic classification - Google Patents

Abstract

The invention discloses a multistage reserve coordination method based on set dynamic classification, and belongs to the technical field of electric power distribution. According to the method, a tendency reserve set is used for guaranteeing consistency of planned output and load change tendency, then, a rapid reserve set is used for preventing supply and demand unbalance caused by predicted deviations of loads, and finally, a real-time reserve set is used for flexibly changing generated output in real time when super-short term fluctuation of the loads happens; and different reserve sets are used sequentially for load adjustment, so that stable and reliable power supply is realized finally. The different reserve sets perform respective duties and support each other on the premise that respective duties are completed, self-positioning is adjusted appropriately, and the sets assist other types of sets to complete the capacity adjustment task.

Description

A kind of coordination approach multistage for subsequent use based on unit dynamic cataloging

Technical field

The invention belongs to distributing electric power technical field, particularly relate to a kind of coordination approach multistage for subsequent use based on unit dynamic cataloging.

Background technology

Electric power have cannot large scale storage, send out the instantaneous feature completed of transmission & distribution, and electric load is along with time, consumption habit, climatic influences and constantly fluctuating, for ensureing the reliability of supply of electric power, grid company needs real time coordination electricity power enterprise to adjust generated output.Under traditional distribution mode, only will ensure that power supply and demand is as Consideration, takes identical scheduling means for different unscheduled case, ignore the efficiency of coordinated scheduling and the requirement of low-carbon environment-friendly.

Summary of the invention

For the traditional distribution mode mentioned in background technology Problems existing in the efficiency and low-carbon environment-friendly of coordinated scheduling, the present invention proposes a kind of coordination approach multistage for subsequent use based on unit dynamic cataloging.

Based on a coordination approach multistage for subsequent use for unit dynamic cataloging, it is characterized in that, described method specifically comprises the following steps:

Step 1: load prediction and forecast: based on the historical data in past three days, social power load situation of change in following 30 minutes of real-time rolling forecast, and the Load Regulation demand of following 15 minutes, 5 minutes and 1 minute three time period is determined according to load variations situation;

Step 2: computer-assisted classification: is divided into 15 minutes the time of power scheduling, 5 minutes and 1 minute three grades; Simultaneously, according to the needs of load variations in time scale and unit reaction speed, machine set type dynamically being divided, is trend unit for subsequent use, quick unit for subsequent use and real-time unit for subsequent use by dispatcher component, determines the reserve capacity of three classes unit for subsequent use and corresponding machine set type;

Step 3: rolling forecast and dynamic conditioning: along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast; Along with load dynamic change, the realize load Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment and the harmonious of reserve capacity; If deficiency for subsequent use, between first existing unit, carry out dynamic conditioning, if still not enough, provide alert, transfer to artificial treatment.

In described step 1, based on the historical data in past three days, the process of the social power load situation of change in following 30 minutes of real-time rolling forecast is:

Take current time as benchmark, try to achieve the load mean value of following 0-30 minute corresponding with first three day respectively:

Y _i=(Y _i,1+Y _i,2+Y _i,3)/3,

Wherein, i=0,5,10,15,20,25,30 are respectively the time point got in following 0-30 minute; Y _{i, 1}be the historical load of this time point before 1 day, Y _{i, 2}be the historical load of this time point before 2 days, Y _{i, 3}it is the historical load of this time point before 3 days;

Actual load D this moment ₀represent, following 5-30 minute required predicted load, uses D respectively ₅, D ₁₀, D ₁₅, D ₂₀, D ₂₅, D ₃₀represent:

D ₅=(D ₀/Y ₀)*Y ₅；

D ₁₀=(D ₀/Y ₀)*Y ₁₀；

D ₁₅=(D ₀/Y ₀)*Y ₁₅；

D ₂₀=(D ₀/Y ₀)*Y ₂₀；

D ₂₅=(D ₀/Y ₀)*Y ₂₅；

D ₃₀=(D ₀/Y ₀)*Y ₃₀。

In described step 1, determine that according to load variations situation the process of the Load Regulation demand of 15 minutes, 5 minutes future and 1 minute three time point is:

1) 15 minutes Load Regulation demand D _qfor the difference that the predicted load of following 15 minutes and generation schedule are exerted oneself, that is:

D _q=Q _f-Q _p

Wherein, Q _ffor predicted load, Q _pfor plan is exerted oneself; D _qextra duty demand is indicated, D during >0 _qindicate during <0 that providing auxiliary power plan is exerted oneself;

2) 5 minutes Load Regulation demand D _kaverage for same period load prediction of front N day deviation:

$D_k=\frac{1}{N}\underset{l=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}_l$

Wherein, e _lbe this moment load prediction deviation of l day, i.e. the difference of load actual value and predicted load; D _krepresent during >0 that prediction data is less than normal in earlier stage, D _krepresent during <0 that prediction data is bigger than normal in earlier stage;

3) 1 minute Load Regulation demand D _sfor load is from the undulating value in 1 minute, the peak load namely in a minute and the difference of minimum load:

D _s=D _m,Max-D _m,Min

Wherein, D _{m, Max}, D _{m, Min}represent the peak load and minimum load that within m minute, occur respectively.

In described step 2, trend unit for subsequent use, quick unit for subsequent use and the machine set type corresponding to real-time unit for subsequent use are respectively:

Machine set type corresponding to described trend unit for subsequent use is: 600MW and above fired power generating unit;

Machine set type corresponding to described quick unit for subsequent use is: Hydropower Unit and 600MW and following fired power generating unit;

Machine set type corresponding to described real-time unit for subsequent use is: automatic-generation-control unit AGC.

In described step 2, determine that the process of the reserve capacity of three classes unit for subsequent use is:

1) for trend unit for subsequent use, the computing formula of a jth unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jqs}}={\mathrm{MaxQ}}_{\mathrm{jq}}-Q_{\mathrm{jqp}}\\ Q_{\mathrm{jqx}}=Q_{\mathrm{jqp}}-{\mathrm{MinQ}}_{\mathrm{jq}}\end{array}\right.$

Wherein: Q _jqs, Q _jqxrepresent the upper capacitance-adjustable of jth trend unit for subsequent use and lower capacitance-adjustable respectively; Q _jqprepresent that the unit plan for subsequent use of a jth trend is exerted oneself; MaxQ _jqrepresent that the maximum technology of jth trend unit for subsequent use is exerted oneself, MinQ _jqrepresent that jth trend unit minimum technology for subsequent use is exerted oneself; The reserve capacity scope of Hydropower Unit is [0, Hydropower Unit rated output]; The reserve capacity scope of fired power generating unit is [50% fired power generating unit rated output, 100% fired power generating unit rated output];

2) computing formula of quick unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jks}}={\mathrm{MaxQ}}_{\mathrm{jk}}-Q_{\mathrm{jkp}}\\ Q_{\mathrm{jkx}}=Q_{\mathrm{jkp}}-{\mathrm{MinQ}}_{\mathrm{jk}}\end{array}\right.$

Wherein, Q _jks, Q _jkxrepresent the upper capacitance-adjustable of a jth quick unit for subsequent use and lower capacitance-adjustable respectively; Q _jkprepresent that a jth quick unit plan for subsequent use is exerted oneself; MaxQ _jkrepresent that a jth maximum technology of quick unit for subsequent use is exerted oneself, MinQ _jkrepresent that a jth quick unit minimum technology for subsequent use is exerted oneself;

3) computing formula of real-time unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jss}}={\mathrm{MaxQ}}_{\mathrm{js}}-Q_{\mathrm{jsp}}\\ Q_{\mathrm{jsx}}=Q_{\mathrm{isp}}-{\mathrm{MinQ}}_{\mathrm{js}}\end{array}\right.$

Wherein, Q _jss, Q _jsxrepresent the upper capacitance-adjustable of a jth real-time unit for subsequent use and lower capacitance-adjustable respectively; Q _jsprepresent that a jth real-time unit plan for subsequent use is exerted oneself; MaxQ _jsrepresent that a jth maximum technology of real-time unit for subsequent use is exerted oneself, MinQ _jsrepresent that a jth real-time unit minimum technology for subsequent use is exerted oneself;

Trend unit reserve capacity >=15 for subsequent use minute Load Regulation demand;

Quick unit reserve capacity >=5 for subsequent use minute Load Regulation demand;

Real-time unit reserve capacity >=1 for subsequent use minute Load Regulation demand.

In described step 3, along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast, and the Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment that realizes load and the harmonious process of reserve capacity are:

1) for the adjustment of trend unit for subsequent use

The adjustment of trend unit for subsequent use meets for the load trend variation adjustment in 15 minutes:

Work as D _qduring >0, if ∑ Q _jqs>=D _q, then unit generated output for subsequent use for trend is improved D _q; Actual power is exerted oneself and equals predicted load; Otherwise, to other province power purchases;

Work as D _qduring <0, if ∑ Q _jqx>=D _q, then unit generated output for subsequent use for trend is reduced | D _q|, actual power is exerted oneself and equals predicted load; Otherwise, to other province sales of electricity;

By trend adjustment for subsequent use, the trend that effectively can realize workload demand and generated output balances.But because load prediction exists certain deviation, therefore, along with the propelling of time, the ultra-short term that 0-15 divides predicts the outcome more accurate, certain deviation may be still there is in new load prediction data with the generated output after trend unit for subsequent use adjusts, now, just need to introduce quick unit for subsequent use to adjust further;

2) quick unit adjustment for subsequent use

At present, the load prediction precision of grid company remains on about 99.3%, still there is certain deviation.Quick adjustment for subsequent use is the adjustment of exerting oneself taked for 0-15 point of internal cause reply load prediction deviation:

Work as D _k>0, if ∑ Q _jks>=D _k, then quick unit generated output for subsequent use is improved D _k, make up load prediction deviation; Otherwise temporarily transfer generating capacity to trend unit for subsequent use, maximum secondment demand is D _k-∑ Q _jks, trend unit for subsequent use, under the prerequisite ensureing self-regulation task, provides adjustment capacity, if trend unit for subsequent use is all in maximum output state, without spare capacity, then waits for the further adjustment of real-time unit for subsequent use;

Work as D _kduring <0, if ∑ Q _jkx>=D _k, then quick unit generated output for subsequent use is reduced | D _k|; Otherwise, to other province sales of electricity;

3) real-time unit adjustment for subsequent use

By above-mentioned two-stage unit adjustment, generated output and workload demand kept in balance, but due to uncontrollable factors such as user power utilization custom, weather conditions, may still there is certain fluctuation in load variations, this kind of fluctuation is predictable low, therefore, by real-time unit for subsequent use---AGC unit adjusts.

AGC is the closed-loop control system utilizing scheduling monitor computer, passage, terminal unit, execution (distribution) device, generating set automation device etc. to form, and the frequency of monitoring, adjustment electric power system, to control generator output.Drop into the unit of AGC-AUTO mode operation, be responsible for tracking prediction deviation and area control error ACE, exerting oneself of automatic frequency adjustment unit.

The invention has the beneficial effects as follows, Each performs its own functions for different unit for subsequent use, under the prerequisite having ensured self task, supports each other, suitably adjust self poisoning, and auxiliary other types unit completes capacity adjustment task.

Accompanying drawing explanation

Fig. 1 is conventional scheduling method flow chart provided by the invention;

Fig. 2 is multistage coordination optimization scheduling flow figure for subsequent use provided by the invention;

Fig. 3 is multistage coordination optimization dispatching method time adjustment figure for subsequent use provided by the invention.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.It should be emphasized that following explanation is only exemplary, instead of in order to limit the scope of the invention and apply.

Fig. 1 is conventional scheduling method flow chart provided by the invention.In Fig. 1, under traditional distribution mode, if load is higher than generation schedule, the load had more can be distributed to standby generator sets according to certain method by control centre.Conventional distribution method has:

(1) different units is given by sharing of load pro rata according to unit installed capacity;

(2) priority allocation is to clean energy resourcies such as water power, and being secondly efficient large fired power generating unit, is common fired power generating unit again.

Under traditional distribution mode, within every 5 minutes, perform a power scheduling, and only have this level, cannot realize dispatching in advance, the wasting of resources of unit for subsequent use may be caused, increase the working strength of dispatcher, meanwhile, power generation dispatching precision may be reduced.

Fig. 2 is multistage coordination optimization scheduling flow figure for subsequent use provided by the invention.Described method specifically comprises the following steps:

Step 1: load prediction and forecast: based on the historical data in past three days, social power load situation of change in following 30 minutes of real-time rolling forecast, and the Load Regulation demand of following 15 minutes, 5 minutes and 1 minute three time period is determined according to load variations situation;

Based on the historical data in past three days, the process of the social power load situation of change in following 30 minutes of real-time rolling forecast is:

Take current time as benchmark, try to achieve the load mean value of following 0-30 minute corresponding with first three day respectively:

Y _i=(Y _i,1+Y _i,2+Y _i,3)/3,

Wherein, i=0,5,10,15,20,25,30 are respectively the time point got in following 0-30 minute; Y _{i, 1}be the historical load of this time point before 1 day, Y _{i, 2}be the historical load of this time point before 2 days, Y _{i, 3}it is the historical load of this time point before 3 days;

Actual load D this moment ₀represent, following 5-30 minute required predicted load, uses D respectively ₅, D ₁₀, D ₁₅, D ₂₀, D ₂₅, D ₃₀represent:

D ₅=(D ₀/Y ₀)*Y ₅；

D ₁₀=(D ₀/Y ₀)*Y ₁₀；

D ₁₅=(D ₀/Y ₀)*Y ₁₅；

D ₂₀=(D ₀/Y ₀)*Y ₂₀；

D ₂₅=(D ₀/Y ₀)*Y ₂₅；

D ₃₀=(D ₀/Y ₀)*Y ₃₀；

Determine that according to load variations situation the process of the Load Regulation demand of 15 minutes, 5 minutes future and 1 minute three time point is:

1) 15 minutes Load Regulation demand D _qfor the difference that the predicted load of following 15 minutes and generation schedule are exerted oneself, that is:

D _q=Q _f-Q _p

Wherein, Q _ffor predicted load, Q _pfor plan is exerted oneself; D _qextra duty demand is indicated, D during >0 _qindicate during <0 that providing auxiliary power plan is exerted oneself;

2) 5 minutes Load Regulation demand D _kaverage for same period load prediction of front N day deviation:

$D_k=\frac{1}{N}\underset{l=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}_l$

Wherein, e _lbe this moment load prediction deviation of l day, i.e. the difference of load actual value and predicted load; D _krepresent during >0 that prediction data is less than normal in earlier stage, D _krepresent during <0 that prediction data is bigger than normal in earlier stage;

3) 1 minute Load Regulation demand D _sfor load is from the undulating value in 1 minute, the peak load namely in a minute and the difference of minimum load:

D _s=D _m,Max-D _m,Min

Wherein, D _{m, Max}, D _{m, Min}represent the peak load and minimum load that within m minute, occur respectively;

Step 2: computer-assisted classification: is divided into 15 minutes the time of power scheduling, 5 minutes and 1 minute three grades; Simultaneously, according to the needs of load variations in time scale and unit reaction speed, machine set type dynamically being divided, is trend unit for subsequent use, quick unit for subsequent use and real-time unit for subsequent use by dispatcher component, determines the reserve capacity of three classes unit for subsequent use and corresponding machine set type;

Trend unit for subsequent use, quick unit for subsequent use and the machine set type corresponding to real-time unit for subsequent use are respectively:

Machine set type corresponding to described trend unit for subsequent use is: 600MW and above fired power generating unit;

Machine set type corresponding to described quick unit for subsequent use is: Hydropower Unit and 600MW and following fired power generating unit;

Machine set type corresponding to described real-time unit for subsequent use is: automatic-generation-control unit AGC;

Determine that the process of the reserve capacity of three classes unit for subsequent use is:

1) for trend unit for subsequent use, the computing formula of a jth unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jqs}}={\mathrm{MaxQ}}_{\mathrm{jq}}-Q_{\mathrm{jqp}}\\ Q_{\mathrm{jqx}}=Q_{\mathrm{jqp}}-{\mathrm{MinQ}}_{\mathrm{jq}}\end{array}\right.$

Wherein: Q _jqs, Q _jqxrepresent the upper capacitance-adjustable of jth trend unit for subsequent use and lower capacitance-adjustable respectively; Q _jqprepresent that the unit plan for subsequent use of a jth trend is exerted oneself; MaxQ _jqrepresent that the maximum technology of jth trend unit for subsequent use is exerted oneself, MinQ _jqrepresent that jth trend unit minimum technology for subsequent use is exerted oneself; The reserve capacity scope of Hydropower Unit is [0, Hydropower Unit rated output]; The reserve capacity scope of fired power generating unit is [50% fired power generating unit rated output, 100% fired power generating unit rated output];

2) computing formula of quick unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jks}}={\mathrm{MaxQ}}_{\mathrm{jk}}-Q_{\mathrm{jkp}}\\ Q_{\mathrm{jkx}}=Q_{\mathrm{jkp}}-{\mathrm{MinQ}}_{\mathrm{jk}}\end{array}\right.$

Wherein, Q _jks, Q _jkxrepresent the upper capacitance-adjustable of a jth quick unit for subsequent use and lower capacitance-adjustable respectively; Q _jkprepresent that a jth quick unit plan for subsequent use is exerted oneself; MaxQ _jkrepresent that a jth maximum technology of quick unit for subsequent use is exerted oneself, MinQ _jkrepresent that a jth quick unit minimum technology for subsequent use is exerted oneself;

3) computing formula of real-time unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jss}}={\mathrm{MaxQ}}_{\mathrm{js}}-Q_{\mathrm{jsp}}\\ Q_{\mathrm{jsx}}=Q_{\mathrm{isp}}-{\mathrm{MinQ}}_{\mathrm{js}}\end{array}\right.$

Wherein, Q _jss, Q _jsxrepresent the upper capacitance-adjustable of a jth real-time unit for subsequent use and lower capacitance-adjustable respectively; Q _jsprepresent that a jth real-time unit plan for subsequent use is exerted oneself; MaxQ _jsrepresent that a jth maximum technology of real-time unit for subsequent use is exerted oneself, MinQ _jsrepresent that a jth real-time unit minimum technology for subsequent use is exerted oneself;

Trend unit reserve capacity >=15 for subsequent use minute Load Regulation demand;

Quick unit reserve capacity >=5 for subsequent use minute Load Regulation demand;

Real-time unit reserve capacity >=1 for subsequent use minute Load Regulation demand.

Step 3: rolling forecast and dynamic conditioning: along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast; Along with load dynamic change, the realize load Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment and the harmonious of reserve capacity; If deficiency for subsequent use, between first existing unit, carry out dynamic conditioning, if still not enough, provide alert, transfer to artificial treatment.

Along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast, and the Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment that realizes load and the harmonious process of reserve capacity are:

1) for the adjustment of trend unit for subsequent use

The adjustment of trend unit for subsequent use meets for the load trend variation adjustment in 15 minutes:

Work as D _qduring >0, if ∑ Q _jqs>=D _q, then unit generated output for subsequent use for trend is improved | D _q|; Actual power is exerted oneself and equals predicted load; Otherwise, to other province power purchases;

Work as D _qduring <0, if ∑ Q _jqx>=D _q, then unit generated output for subsequent use for trend is reduced | D _q|, actual power is exerted oneself and equals predicted load; Otherwise, to other province sales of electricity;

By trend adjustment for subsequent use, the trend that effectively can realize workload demand and generated output balances.But because load prediction exists certain deviation, therefore, along with the propelling of time, the ultra-short term that 0-15 divides predicts the outcome more accurate, certain deviation may be still there is in new load prediction data with the generated output after trend unit for subsequent use adjusts, now, just need to introduce quick unit for subsequent use to adjust further.

2) quick unit adjustment for subsequent use

At present, the load prediction precision of grid company remains on about 99.3%, still there is certain deviation.Quick adjustment for subsequent use is the adjustment of exerting oneself taked for 0-15 point of internal cause reply load prediction deviation:

Work as D _k>0, if ∑ Q _jks>=D _k, then quick unit generated output for subsequent use is improved D _k, make up load prediction deviation; Otherwise temporarily transfer generating capacity to trend unit for subsequent use, maximum secondment demand is D _k-∑ Q _jks, trend unit for subsequent use, under the prerequisite ensureing self-regulation task, provides adjustment capacity, if trend unit for subsequent use is all in maximum output state, without spare capacity, then waits for the further adjustment of real-time unit for subsequent use;

Work as D _kduring <0, if ∑ Q _jkx>=D _k, then quick unit generated output for subsequent use is reduced | D _k|; Otherwise, to other province sales of electricity;

3) real-time unit adjustment for subsequent use

By above-mentioned two-stage unit adjustment, generated output and workload demand kept in balance, but due to uncontrollable factors such as user power utilization custom, weather conditions, may still there is certain fluctuation in load variations, this kind of fluctuation is predictable low, therefore, by real-time unit for subsequent use---AGC unit adjusts.

AGC is the closed-loop control system utilizing scheduling monitor computer, passage, terminal unit, execution (distribution) device, generating set automation device etc. to form, and the frequency of monitoring, adjustment electric power system, to control generator output.Drop into the unit of AGC-AUTO mode operation, be responsible for tracking prediction deviation and area control error ACE, exerting oneself of automatic frequency adjustment unit.

Fig. 3 is multilevel coordination Optimization Scheduling time adjustment figure provided by the invention.In Fig. 3, for time point A, the actual load level and the generated output that finally realize A point if think are consistent, then within 15 minutes, call trend unit for subsequent use in advance, plan is exerted oneself and to be consistent with the prediction load of A point; As time goes on, when time left 5 minutes, may there is deviation with the data of 15 minutes in the load data that now A point upgrades, then call quick unit for subsequent use, load deviation part eliminated; When only having 1 minute close to A point, call real-time unit for subsequent use, eliminate accident at any time to the impact of load.

The above; be only the present invention's preferably embodiment, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; the change that can expect easily or replacement, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claim.

Claims (2)

1. based on a coordination approach multistage for subsequent use for unit dynamic cataloging, it is characterized in that, described method specifically comprises the following steps:

Step 1: load prediction and forecast: based on the historical data in past three days, social power load situation of change in following 30 minutes of real-time rolling forecast, and the Load Regulation demand of following 15 minutes, 5 minutes and 1 minute three time period is determined according to load variations situation;

Step 2: computer-assisted classification: is divided into 15 minutes the time of power scheduling, 5 minutes and 1 minute three grades; Simultaneously, according to the needs of load variations in time scale and unit reaction speed, machine set type dynamically being divided, is trend unit for subsequent use, quick unit for subsequent use and real-time unit for subsequent use by dispatcher component, determines the reserve capacity of three classes unit for subsequent use and corresponding machine set type;

Step 3: rolling forecast and dynamic conditioning: along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast; Along with load dynamic change, the realize load Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment and the harmonious of reserve capacity; If deficiency for subsequent use, between first existing unit, carry out dynamic conditioning, if still not enough, provide alert, transfer to artificial treatment;

In described step 1, determine that according to load variations situation the process of the Load Regulation demand of 15 minutes, 5 minutes future and 1 minute three time point is:

1) 15 minutes Load Regulation demand D _qfor the difference that the predicted load of following 15 minutes and generation schedule are exerted oneself, that is:

D _q＝Q _f-Q _p

Wherein, Q _ffor predicted load, Q _pfor plan is exerted oneself; D _qextra duty demand is indicated, D during >0 _qindicate during <0 that providing auxiliary power plan is exerted oneself;

2) 5 minutes Load Regulation demand D _kaverage for same period load prediction of front N day deviation:

$D_k=\frac{1}{N}\underset{l=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}_l$

Wherein, e _lbe this moment load prediction deviation of l day, i.e. the difference of load actual value and predicted load; D _krepresent during >0 that prediction data is less than normal in earlier stage, D _krepresent during <0 that prediction data is bigger than normal in earlier stage;

3) 1 minute Load Regulation demand D _sfor load is from the undulating value in 1 minute, the peak load namely in a minute and the difference of minimum load:

D _s＝D _m,Max-D _m,Min

Wherein, D _{m, Max}, D _{m, Min}represent the peak load and minimum load that within m minute, occur respectively;

In described step 2, trend unit for subsequent use, quick unit for subsequent use and the machine set type corresponding to real-time unit for subsequent use are respectively:

Machine set type corresponding to described trend unit for subsequent use is: 600MW and above fired power generating unit;

Machine set type corresponding to described quick unit for subsequent use is: Hydropower Unit and 600MW and following fired power generating unit;

Machine set type corresponding to described real-time unit for subsequent use is: automatic-generation-control unit AGC;

In described step 2, determine that the process of the reserve capacity of three classes unit for subsequent use is:

1) for trend unit for subsequent use, the computing formula of a jth unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jqs}}=\mathrm{Max}{Q}_{\mathrm{jq}}-Q_{\mathrm{jqp}}\\ Q_{\mathrm{jqx}}=Q_{\mathrm{jqp}}-\mathrm{Min}{Q}_{\mathrm{jq}}\end{array}\right.$

Wherein: Q _jqs, Q _jqxrepresent the upper capacitance-adjustable of jth trend unit for subsequent use and lower capacitance-adjustable respectively; Q _jqprepresent that the unit plan for subsequent use of a jth trend is exerted oneself; MaxQ _jqrepresent that the maximum technology of jth trend unit for subsequent use is exerted oneself, MinQ _jqrepresent that jth trend unit minimum technology for subsequent use is exerted oneself; The reserve capacity scope of Hydropower Unit is [0, Hydropower Unit rated output]; The reserve capacity scope of fired power generating unit is [50% fired power generating unit rated output, 100% fired power generating unit rated output];

2) computing formula of quick unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jks}}=\mathrm{Max}{Q}_{\mathrm{jk}}-Q_{\mathrm{jkp}}\\ Q_{\mathrm{jkx}}=Q_{\mathrm{jkp}}-\mathrm{Min}{Q}_{\mathrm{jk}}\end{array}\right.$

Wherein, Q _jks, Q _jkxrepresent the upper capacitance-adjustable of a jth quick unit for subsequent use and lower capacitance-adjustable respectively; Q _jkprepresent that a jth quick unit plan for subsequent use is exerted oneself; MaxQ _jkrepresent that a jth maximum technology of quick unit for subsequent use is exerted oneself, MinQ _jkrepresent that a jth quick unit minimum technology for subsequent use is exerted oneself;

3) computing formula of real-time unit reserve capacity for subsequent use is:

$\left\{\begin{array}{c}{Q}_{\mathrm{jss}}=\mathrm{Max}{Q}_{\mathrm{js}}-Q_{\mathrm{jsp}}\\ Q_{\mathrm{jsx}}=Q_{\mathrm{isp}}-\mathrm{Min}{Q}_{\mathrm{js}}\end{array}\right.$

Wherein, Q _jss, Q _jsxrepresent the upper capacitance-adjustable of a jth real-time unit for subsequent use and lower capacitance-adjustable respectively; Q _jsprepresent that a jth real-time unit plan for subsequent use is exerted oneself; MaxQ _jsrepresent that a jth maximum technology of real-time unit for subsequent use is exerted oneself, MinQ _jsrepresent that a jth real-time unit minimum technology for subsequent use is exerted oneself;

Trend unit reserve capacity >=15 for subsequent use minute Load Regulation demand;

Quick unit reserve capacity >=5 for subsequent use minute Load Regulation demand;

Real-time unit reserve capacity >=1 for subsequent use minute Load Regulation demand;

In described step 3, along with the propelling of time, trend of calling successively is for subsequent use, for subsequent use and real-time unit for subsequent use fast, and the Load Regulation demand in 15 minutes, 5 minutes, 1 minute three grades of moment that realizes load and the harmonious process of reserve capacity are:

1) for the adjustment of trend unit for subsequent use

The adjustment of trend unit for subsequent use meets for the load trend variation adjustment in 15 minutes:

Work as D _qduring >0, if ∑ Q _jqs>=D _q, then unit generated output for subsequent use for trend is improved D _q; Actual power is exerted oneself and equals predicted load; Otherwise, to other province power purchases;

Work as D _qduring <0, if ∑ Q _jqx>=D _q, then unit generated output for subsequent use for trend is reduced | D _q|, actual power is exerted oneself and equals predicted load; Otherwise, to other province sales of electricity;

2) quick unit adjustment for subsequent use

Work as D _k>0, if ∑ Q _jks>=D _k, then quick unit generated output for subsequent use is improved D _k, make up load prediction deviation; Otherwise temporarily transfer generating capacity to trend unit for subsequent use, maximum secondment demand is D _k-∑ Q _jks, trend unit for subsequent use, under the prerequisite ensureing self-regulation task, provides adjustment capacity, if trend unit for subsequent use is all in maximum output state, without spare capacity, then waits for the further adjustment of real-time unit for subsequent use;

Work as D _kduring <0, if ∑ Q _jkx>=D _k, then quick unit generated output for subsequent use is reduced | D _k|; Otherwise, to other province sales of electricity;

3) real-time unit adjustment for subsequent use

Drop into the unit of AGC-AUTO mode operation, be responsible for tracking prediction deviation and area control error ACE, exerting oneself of automatic frequency adjustment unit.

2. method according to claim 1, is characterized in that, in described step 1, based on the historical data in past three days, the process of the social power load situation of change in following 30 minutes of real-time rolling forecast is:

Take current time as benchmark, try to achieve the load mean value of following 0-30 minute corresponding with first three day respectively:

Y _i＝(Y _i,1+Y _i,2+Y _i,3)/3；

Wherein, i=0,5,10,15,20,25,30 are respectively the time point got in following 0-30 minute; Y _{i, 1}be the historical load of this time point before 1 day, Y _{i, 2}be the historical load of this time point before 2 days, Y _{i, 3}it is the historical load of this time point before 3 days;

Actual load D this moment ₀represent, following 5-30 minute required predicted load, uses D respectively ₅, D ₁₀, D ₁₅, D ₂₀, D ₂₅, D ₃₀represent:

D ₅＝(D ₀/Y ₀)×Y ₅；

D ₁₀＝(D ₀/Y ₀)×Y ₁₀；

D ₁₅＝(D ₀/Y ₀)×Y ₁₅；

D ₂₀＝(D ₀/Y ₀₎×Y ₂₀；

D ₂₅＝(D ₀/Y ₀)×Y ₂₅；

D ₃₀＝(D ₀/Y ₀)×Y ₃₀。